I have always been fascinated by the ability of electrical components like resistors, capacitors and inductors to be connected in not just one way but two ways, in series and in parallel. But its not until recently that I have looked into the underlying formulas to determine how to utilize it in a favorable fashion energy-wise, yes I'm that lazy.

But before I present my idea I would like to elaborate on what the effect might be if there was two ways of adding masses into one, the “ordinary” way m=m1+m2, and a so far “unknown” way m=1/(1/m1+1/m2).

The formulas I will use are E=½*m*v² for kinetic energy and p=m*v for momentum.

Assume two masses with equal mass m. Accelerate them when they are added in the “ordinary” way to 2*m. When the acceleration stops they will have speed v, kinetic energy will be E=½*2*m*v² and momentum will be p=2*m*v. Now, rearrange the masses so that they are added in the “unknown” way to ½*m and assume Newton was right in that momentum will always be preserved. This means that to preserve momentum speed has to change like a departing UFO from v to 4*v so that p=2*m*v=½*m*4*v. If the new speed is 4*v the new energy is E=½*½*m*(4*v)²=4*m*v² which is 4 times higher than when the masses are added in the “ordinary” way.

COP=4.

Now lets do this with coils. Energy in a coil is E=½*L*I² and the entity corresponding to momentum would be the magnetic flux which is Φ=L*I. A battery, a switch and two windings on the same core are in series. Energize the coil and turn the switch off. Collect the flyback with a diode but just from *one* of the windings (or from both if they are rearranged and combined in parallel). If momentum is preserved, that is, if magnetic flux is preserved, the energy in the flyback is 4 times the input (minus losses) as above.

I have tried this without success so far, but my switch has either been a 2N3055 transistor which I guess is too slow, or a relay that just generates sparks and get stuck.

A spark-gap and a step-down transformer would probably do the trick. It seems like most OU-probable circuits have this in them somewhere.

So I tried another approach. Using the “stingo”-circuit (2N3055/2N2955) to energize a bifilar coil with its windings in series and capturing the flyback in a capacitor with a resistor in parallel as in the first pic I measured the voltage over the capacitor and resistor to 17.1 V. Then I connected one capacitor with a resistor on *each* winding as in the second picture, same type/size of capacitor and resistor and diode for each winding. This has the feeling of the two windings in parallel. I trimmed the circuit slightly to make it draw the same amount as in the first setup. The voltage over each capacitor and resistor was now 11.9 V. This is a slight gain in energy. (Edit: no its not, ill try new setups though...)

/Hob

Attached Images

	1-cap.png (936 Bytes, 59 views)
	2-cap.png (1.1 KB, 49 views)

Your reasoning is very good. A while ago I made a thread about a similar concept but which is mechanical.

Create energy with conservation of momentum law.

It makes sense that the flux remains constant so changing L should change I.

Yes, same thing, in my case i need to switch off the current as abrupt as possible to mimic something of an reverse elastic-collision situation.

/Hob

I think you should search Jerry Voland work too, since he obtain some interesting phenomena with spark. Unfortunately, still clue, not solution.

MIT doing it with a capacitor?

The energy formula is E=½*C*U²
and the "momentum" formula is Q=C*U

When the charge Q is constant
and capacitance C drops
the voltage U rise.
What happens to the stored energy?

YouTube - MIT Physics Demo -- Adjustable Capacitor with Dielectric

/Hob

@nilrehob: there's something we forgot. The parameter we adjust, windage, in order to change inductance has actually a squared relationship in our inductance equation and a linear one in flux. Flux isn't L*I but L*I/N. Here are some relevant equations for a straight solenoid:



That means if we for instance halve the amount of windings the current would double so flux is conserved. However our inductance will actually drop 4 times, not 2. Compensating for the increased square amperage in the energy equation.

In other words the windings do not have a linear relationship with mass. But a squared one. So this defeats our point. However there are other parameters that do have this linear relation. Area and permeability. If you can change those during operation the change would directly be reflected in an equal change of current through conservation of flux. And since they have a linear relation in both the flux and energy equation you will thus see an increase in inductive energy.

@broli

Yes, youre right, the inductance is not added as H+H when sharing core.
This is however only to our benefit,
as theoretical COP is Hin/Hout. (1)

This comes from COP = Eout/Ein = ½*Hout*Iout² / ½*Hin*Iin² (2)
since (if) flux is constant Hout*Iout = Hin*Iin (3)
from (3) you have Iout = Hin*Iin / Hout (4)
put that into (2) and you have ½*Hout*(Hin*Iin / Hout)² / ½*Hin*Iin² (5)
simplify this and you get Hin/Hout (1)

But maybe I need to digest your point some more.

/Hob

@broli

So if you have a 6-filar
put 2 winds in series, thats for input
and 4 winds in parallel, for output
might do the trick?

/Hob

The problem will persist. As you said, digest more what I said. It's a mathematical problem. When you decrease windings 2, 3 or 4 times the current will indeed rise 2, 3 and 4 times. However energy is dependent on windings squared times current squared. So mathematically you can write:



From conservation of flux we know that the product N*I remains constant. So that means energy will remain constant as well no matter the combination.

@broli

Doh!

So lets use a toroid with a ferrite core, all flux is now in the core, in the Weiss-domains of the ferrite,
now this has to work, please...

BTW, doesn't ferrite have a kind of delay for getting magnetized and demagnetized? Whats it called? If there is one wouldn't it dictate the maximum switch-off-time?

/Hob

Well, tomorrow I plan to make a toroidal transformer.
I don't have a proper or fancy core so I'm going to use a plastic-coated "garden-wire" which has some iron in it it seems.
Maybe I shall do the core as a torus instead, with the coil-wires inside?
First wind the coil-wires like an air-core coil, and then wind the garden-wire around it?

/Hob

When you use a core you change the game, because then you start to play with permeability. Permeability change would truly change energy because it has both a linear relation with flux and energy. But the problem is that we can't instantly switch between permeabilities like we could between winding amount...or can we? For instance a ferromagnetic material losses its magnetic properties when it's heated above a point. You will gain energy through conservation of flux, the heat will go nowhere, so in theory you can extract the increased inductive energy, recapture the heat and start all over. But using heat would be terribly inefficient.

Different layers of coil-wire and garden-wire will change the permeability.
I guess the coil-part with the smaller induction H shall have the greater amount of core?
Or do I have it backwards again?

/Hob

Does H stand for Henry or the H field?

I meant Henry, sorry for being unspecific.

/Hob

Just analyze the equations carefully. And look at the raw parameters that make up the inductance of a solenoid. For instance I just noticed we completely ignored coil length so far. If you have two coils on top of each other, with the same amount of windings but one longer than the other. When energizing the short one and opening the switch and letting the long one collect the inductive energy what you did is instantly and effortlessly double the length, this would halve the flux, so conservation of flux kicks in and doubles current. Since inductive energy is linearly proportional to 1/length as well you should see an increase.

You're right. Then one solution would be using thicker wire for the secondary and keep the same number of turns (but not wind them together).
So which part of the kapagen do You think we are looking at?

/Hob

That's an interesting last sentence but I don't want to go into the kapagen here as it has enough threads and imo would just dilute the logical train of thoughts this thread started with.

The trick is to try and trap as much flux as possible. That's why a closed loop core like a toroid could be beneficial. It's well known that spacing of the windings of the toroids affects the inductance significantly:

http://users.catchnet.com.au/~rjandu...s/wind_deg.gif

I would suppose the bigger the toroid the more you can space the windings apart and increase the flux difference. If we suppose most of the flux gets trapped. Then switching from a densely wound winding to a spaced out one, will give maximum flux difference and thus shoot the current up to conserve it. This all without changing the amount of windings. So inductive energy has to go up too.

Great pic of the toroid, thanks!

But I must admit there is still an enigma left in the formulas for me.
We have the definition of the unit Henry:

Which gives you Wb=H*A, or Φ=L*I, right?
Then we have the definition of flux:

Combine them and you get L=μ*N*A/l
But that's not right! It should be

Where is the error?



/Hob

flux is not Φ=L*I but it is Φ=L*I/N. Amount of turns has no unit. The reason why you need to divide by N is because the inductance already has amount of windings squared (N^2) incorporated.
It's easier to see this if you derive the "inductance" from Faraday's law (which is where it comes from in the first place):

Yes, I just found it myself
L = N × d Φ / d I
at Magnetism: quantities, units and relationships

/Hob

Now, back to your toroid-pic:

This reminds me very much of the toroid in the vid starting this thread:
This is it !

/Hob

So what about three parallel windings,
each occupying 120deg of the toroid and having N turns each for input,
and then a wider and thicker winding with N turns on top (or under?) as output?
But on the other hand, parallel windings lower the inductance ...

I have to dig up the Bob Boyce documents lying on my hard-drive somewhere.

/Hob

I would wind each consequent winding next to the previous, so basically like a bifilair, trifilair... coil. This in order that 10 windings of the "bigger spaced" coil covers as much area and leaves no gaps without having to actually use a real broad wire. While your input coil has 10 windings which are spaced very closely. This also brings us to a more accurate flux equation for a toroid:




theta*r is really just definition of arc length. In textbooks theta is always equal to 2pi radians which is 360°.

Hi nilrehob and broli, Good to see you guys doing the math on this, the more people that look into this the better. I thought I would give you guys this link to look at. If you look at page 110 figure 92 and 93, also page 111 figure 94.

Page 110 figure 92 and 93 is the Tesla transformer/converter and page 111 figure 94 is the same thing with generator attached. The section "High Frequency Polyphase Transformer" starts on page 109, it's very interesting, in my opinion it can be a "flux flywheel generator" and that is kinda how it's descibed in there.

Tesla used quad primaries with the oposite pairs series connected and fired 90 degrees out of phase, which was actually 180 out of phase on the generator.

I think, maybe you guys can decifer that better than me. He also used quad secondaries and he shows that the secondaries can be connected however is needed or desired, it's a very cool setup. And he also used iron shielding wire aswell sometimes.

The inventions, researches and writings of Nikola Tesla, with special reference to his work in polyphase currents and high potential lighting : Martin, Thomas Commerford, 1856-1924 : Free Download & Streaming : Internet Archive

Enjoy
Cheers

On third thought, parallel windings on the same core doesn't affect inductance, just resistance, right?

/Hob

This is getting so interesting i might just get myself a big fat toroid core, at last!

So it all boils down to a theoretical COP=θin/θout, right?

/Hob

Correct, and it's a way to have a "thicker" wire like you mentioned earlier without really having to use a thicker wire.

So it all boils down to a theoretical COP=θout/θin, right?

Edit: since θout is always 2π COP is proportional to 1/θin, or just 1/θ for short.

/Hob

In absolute theoretical cases. But I doubt it be as significant as that in practice. But it's a rather simple experiment. You can compare inductive energy like you did before by converting it to electrical in a cap. 2 Setups:

Primary small angled coil energized and then letting it discharge into a cap. Then same experiment but this time you put the cap on the big angled coil. And hopefully you'll see more voltage there in the cap.

The other experiment needs you to energy both coils in series and discharge them in series, and comparing it to only opening the switch on the small angled coil and discharging through the second in the same cap.